Exothermic Control

[controlnovice]

We are looking at two different methods of an exothermic batch control for DeltaV. The plant likes the current heat control, but the way it's tuned, its pretty much on/off control.

I am looking at one method, another engineer has another preference... and I'm sure there's many more. Any books out there on Exothermic control? I can find minor details in other control books, but nothing real solid.

Can anyone help distinguish between the following? And which would work better?

1. Cascade/Split Range Control: Master controller looks at Reactor Temperature. Output of Master controller goes to splitter. One side goes to the slave controller for the cooling valve. The other side goes to the slave controller for the heating valve (hot oil). The split will not be 50/50 because the process gain for heating is much higher than cooling, so the plan is somewhere around 0-30% output for cooling and 31-100 for heating (but we can play with it).

The heating slave controller has the Hot Oil temperature input. The cooling valve controller has the cooling water outlet temperature input.

2. Separate controllers: Two seperate controllers. One for the heating valve with the reactor temperature input and one for the cooling valve with reactor temperature input. There is also a 'deadband' controller so when the setpoint is, say 300 F, and the temperature increases to 305, the cooling controller will start to come on, and when the temp drops to 295, the heating controller will come on.

This method, it is said, will be better because since the process gains are different, the tuning parameters of the heating controller and cooling controller can be different and better control can be had using two seperate controllers.

Any thoughts on the two options would greatly be appreciated. Any other ideas would be welcome, but I also like to live under the K.I.S.S. method....

 

[FrancisL]

It would be good to see a P&ID and loop diagrams to understand this better, but:
Solution 2 is not right - you should never have two PID controllers simultaneously controlling the same process variable.
Solution 1 sounds ok, and by tuning the split range (which should allow for both the cross-over point and the gains of the two outputs to be adjusted) you should be able to eliminate the effect of differing process gains.

 

[controlnovice]

FrancisL: As my username suggests, I am a beginner control engineer (not fresh out of school though). I know that more than one controller should not be used to control the same process variable. Are there any 'concrete' rules for that or is it just a 'non-written' rule? Also, the two controllers would be controlling different valves. Is that any different, or are we still looking at the same PV?

 

[FrancisL]

II am sure that there is a theoretical proof, but it's a long time since I did my Degree - in Control Engineering.
Intuitively it is obvious - the two controllers may end up fighting each other and potentially you may end up heating and cooling at the same time.

Another point about your exothermic process. If it has the potential to run away, then you need to detect when this is happening and slam on the cooling as fast as possible.
If this is possible then you may need to have other protection systems in place (relief valves sized correctly etc) otherwise your control loops becoming safety critical. You cannot rely on ‘plain’ DeltaV for safety critical controls.

By the way, I also follow control.com, and I see you have posted there as well, but I will stick with Eng-Tips, it has a much better user interface.

http://www.controldraw.co.uk

 

[Guidoo]

Controlnovice,

Just a couple of remarks:

- From your description of the "2 controller" option I understand that these controllers will have P (=Proportional) action only. The way you describe it it can work since only one controller will be active at any moment of time (cooler controller when temperature > 300F and heating controller when temperature <300F). This system will not work when you have Integral action as well.

- It is not uncommon to have two controllers "controlling" the same PV. This is the case for protective controllers. For example, pressure in a distillation column is controlled by a PC acting on condenser (e.g. air cooler fan speed), and another "high pressure protective" PC that opens valve to flare header in case pressure exceeds its SP. This second PC has P-action only, and is there to prevent activation of high pressure trip, opening of relief valve etc.

- For your system, I don't think you can live with controllers with P action only because you cannot accept offset (PV not equal to SP in steady state) requiring Integral action, and you may need D action as well to compensate for slow temperature measurement. Therefore, I don't think you can use the two controller option. In other words, split range controllers are preferred if you want to have tight temperature control.

- For your split-range option, you write that reactor TC is master controller providing setpoints to slave TC controllers in heating and cooling medium. Prerequisite for the beneficial operation of cascade control is that slave loop is about ten times or more faster than master loop. Having a slow TC loop in cascade with another slow TC loop is normally not such a good idea. Why don't you have the reactor TC loop act directly on the valves (via split range ofcourse)? Do you expect frequent, fast and large variations in heating and/or cooling medium temperature, that can affect the reactor temperature? If that is the case, is it possible to solve that at the source by improving heating/cooling medium temperature control.

- In order to prevent opening of heating and cooling valve at same time I would consider to have a gap in your split range, e.g. 0-25% cooling and 35-100% heating.

- As you wrote yourself, the issue of having different process gains can be tackled by carefully configuration of the split-range.

 

[FrancisL]

Just a quick point re
- It is not uncommon to have two controllers "controlling" the same PV.
I suggest that this applies when one PID overrides the other and that in reality only one is in control at any one time.

 

[Guidoo]

Francis,

Correct, that is why I used the quotation marks "" around "controlling".

 

[controlnovice]

Thanks for the info!!

FrancisL: Yes this is a better format than control.com! I am using a contraint controller (really full cooling on - no control) for the exotherm. I am caclulating the temperature rate of rise, and when it gets above a value, say 5 deg/min, then heat is turned off and cooling is full on.

Guidoo: One engineer is currently using the two controller method. I'm not real thrilled with it. But instead of PID control, he is using fuzzy logic. So I can't answer the P vs I comments.

As far as using Cascade, I am under the impression that the slave loops should be around 3 times faster than the master. I suppose the hot oil temperature will not vary much, so we may not need cascade on the heating side. However, on one product, we need to limit the hot oil temperature to the reactor, so I thought by going to a slave controller on the hot oil side, I could set a maximum setpoint for the hot oil temp. One thing to note about heating, the hot oil is continuously circulating during heating and exotherm, only the hot oil supply to the pump is controlled. During cooling (when no heat is desired), the pump is turned off. So, when we are at exotherm, the pump is still pumping oil, the hot oil valve could be closed, and we could be sending cooling water to the reactor.

On the cooling water side, the cooling water temperature can vary during the heating process, and especially since we have more than one reactor. So that is why the slave is there. However, usually when we want cooling, I suppose we really don't care what the temperature of the water is. We just want it to cool. I mean, we don't want to send water that is too hot back to the tower, but that is secondary to maintaining control of the reactor.

Oh, also about having a gap between the valves (0-25, and 35-100%): I've noticed on other processes that many times no process flows until a valve is open around 3-5%, so I was going to make the gap much closer (2% differerence), or even overlap slightly (2-5% overlap) which may assist due to the pump still pumping oil which is still hot.

Responses would be appreciated!

 

[Guidoo]

As was to be expected, similar systems have been described in control systems literature. For example, in section 8.24 of book "Process Control" by Béla G. Lipták. He recommends to have a recirculated cooling water configuration to ensure a constant and high rate of water circulation. The coolant loop requires a pump and contains a heat exchanger so that the cooling water can be heated (e.g. with steam) in order to get the reaction started. There is a master reaction temperature controller, a primary slave jacket temperature controller and secondary slave valve position controllers on the cooling water and steam supply valves. The split range configuration (between jacket TC and valve position controllers) is as described in previous posts.

Lipták gives several suggestions on how the system should be configured, for example:
- the master TC should have anti-reset windup
- slave TC normally has proportional-plus-derivative or proportional only